Maps

Module 10

Ray J. Hoobler

Libraries

Code

library(tidyverse)
library(patchwork)
library(maps) # Original maps package (mostly outdated, but can still be used for simple maps)
library(usmap) # US map that includes Alaska and Hawaii 
library(mapdata) # Extra map databases (supplment to maps package; world map is out-dated)
library(ggmap) # Allows use of Stamen and Google maps with ggplot2
library(sf) # Simple Features for R (modern way to work with spatial data)

Simple Maps (1/2)

Getting started with the `maps` package

Code

map("state") # US state map (lower 48)

Simple Maps (2/2)

map(database = "world", regions = ".", exact = FALSE, boundary = TRUE,
  interior = TRUE, projection = "", parameters = NULL, orientation = NULL,
  fill = FALSE, col = 1, plot = TRUE, add = FALSE, namesonly = FALSE,
  xlim = NULL, ylim = NULL, wrap = FALSE, resolution = if (plot) 1 else 0,
  type = "l", bg = par("bg"), mar = c(4.1, 4.1, par("mar")[3], 0.1),
  myborder = 0.01, namefield="name", lforce="n", ...)

Using Data from the `maps` Package in `ggplot2` (1/3)

Create a data frame from the `maps` package databases

Code

us_states <- map_data("state")
us_states

Using Data from the `maps` Package in `ggplot2` (2/3)

Create a map of the US states using `ggplot2`

Code

ggplot(us_states, aes(x = long, y = lat, group = group)) + # make sure to include "group" in the aesthetics 
  geom_polygon(fill = "white", color = "black") + # borders on the map are just a collection of polygons; each polygon is a group
  theme_void()

The scaling is off as ggplot does not know how to scale the map.

Using Data from the `maps` Package in `ggplot2` (3/3)

Two solutions for scaling the map

1. Use the `coord_fixed()` function

Code

ggplot(us_states, aes(x = long, y = lat, group = group)) + 
  geom_polygon(fill = "white", color = "black") + 
  theme_void() +
  coord_fixed(1.3)

2. Use the `coord_map()` function

Code

ggplot(us_states, aes(x = long, y = lat, group = group)) + 
  geom_polygon(fill = "white", color = "black") + 
  theme_void() +
  coord_map()

Add Cities to the Map (1/4)

Create a data frame of cities in the US usign the `us.cities` data set

Code

us_cities <- as_tibble(us.cities)
us_cities

Add Cities to the Map (2/4)

Create a map of the US states with cities using `ggplot2`

Code

us_states |> 
  ggplot(aes(x = long, y = lat)) + 
  geom_polygon(aes(group = group), fill = "white", color = "black") + 
  geom_point(data = us_cities, aes(x = long, y = lat), color = "blue", size = 1) +
  theme_void() +
  coord_map()

Add Cities to the Map (3/4)

What’s a quick way to remove Alaska and Hawaii cities from the map?

Code

cities_lower48 <- us_cities |>
  filter(!(country.etc %in% c("AK", "HI")))

us_states |> 
  ggplot(aes(x = long, y = lat)) + 
  geom_polygon(aes(group = group), fill = "white", color = "black") + 
  geom_point(data = cities_lower48, aes(x = long, y = lat), color = "blue", size = 1) +
  theme_void() +
  coord_map()

Add Cities to the Map (4/4)

Add state capitals and the national capital as unique points

Code

state_capitals <- cities_lower48 |>
  filter(capital == 2)

us_capital <- us_cities |>
  filter(name == "WASHINGTON DC")

us_states |> 
  ggplot(aes(x = long, y = lat)) + 
  geom_polygon(aes(group = group), fill = "white", color = "black") + 
  geom_point(data = cities_lower48, aes(x = long, y = lat), color = "blue", size = 1, shape = 1) +
  geom_point(data = state_capitals, aes(x = long, y = lat), color = "red", size = 2) +
  geom_point(data = us_capital, aes(x = long, y = lat), fill = "green", size = 3, shape = 23) +
  theme_void() +
  coord_map()

Map projections

Code

us_states |> 
  ggplot(aes(x = long, y = lat)) + 
  geom_polygon(aes(group = group), fill = "white", color = "black") + 
  geom_point(data = cities_lower48, aes(x = long, y = lat), color = "blue", shape = ".") +
  coord_map("albers", lat0 = 30, lat1 = 45)

Maps that Include Alaska and Hawaii (1/2)

Code

usmap::plot_usmap(regions = "states")

Maps that Include Alaska and Hawaii (2/2)

the `usmap::plot_usmap()` function returns a ggplot object

Code

plot_usmap(regions = "states") +
  labs(title = "US States")

Inspect the `us_map()` Data Frame

Note, we do not have a data frame with latitude and longitude columns

Code

state_df <- us_map()
state_df

Adding Cities to the `usmap` Map

A convenience method called `usmap_transform` can be used to add cities (or other points) to the map.

Code

us_cities_transformed <- usmap_transform(us_cities, input_names = c("long", "lat"))

plot_usmap() +
  geom_sf(data = us_cities_transformed, aes(size = pop, color = log10(pop)))

Choropleth Maps (from Claude.ai)

what is the origin of the word choropleth?

The word “choropleth” has an interesting etymology rooted in Greek. It’s a compound word formed from two Greek elements:

“Choro-” (χώρο-): This prefix comes from the Greek word “choros” (χώρος), which means “area” or “region.”
“-pleth” (-πλήθ): This suffix is derived from the Greek word “plethos” (πλῆθος), meaning “multitude” or “quantity.”

When combined, these elements create “choropleth,” which essentially means “area quantity” or “region value.” This name aptly describes the nature of choropleth maps, which display quantities or values across different geographic areas.

The term “choropleth” was coined in 1938 by John Kirtland Wright, an American geographer. Wright introduced this term in his doctoral dissertation at Harvard University, titled “Notes on Statistical Mapping, With Special Reference to the Mapping of Population Phenomena.”

This relatively recent coinage explains why the term might seem unfamiliar to many people, despite the widespread use of choropleth maps in various fields today.

Add a Choropleth Layer to the Map

State Population Data

Code

plot_usmap(data = statepop, values = "pop_2022", color = "white") +
  scale_fill_continuous(
    low = "black", high = "red",
    name = "Population (2022)",
    label = scales::comma,
    trans = "log10") +
  theme(legend.position = "right")

Code

plot_usmap(data = statepop, values = "pop_2022", color = "white") +
  scale_fill_viridis_c(
    name = "Population (2022)",
    label = scales::comma,
    trans = "log10") +
  theme(legend.position = "right")

Working with Shapefiles

What are shapefiles?

From GitHub Copilot:

Shapefiles are a popular geospatial vector data format developed by Esri. They are used to store the geometric location and attribute information of geographic features. Shapefiles consist of several files with the same name but different extensions, such as .shp, .shx, .dbf, and .prj.

The .shp file contains the geometry data, such as points, lines, or polygons, that represent the geographic features.
The .shx file is an index file that helps speed up access to the geometry data.
The .dbf file stores the attribute data associated with the geographic features, such as names, population counts, or other information.
The .prj file contains the coordinate system information used to project the data onto a map.

US Census

Census Mapping Files

Loaing Shapefiles

Code

# Read the shapefile
ut_shapefile <- st_read("shapefiles/tl_2024_49_tract/tl_2024_49_tract.shp")

Reading layer `tl_2024_49_tract' from data source 
  `/Users/ray/Documents/Documents - RAY’s Mac Studio/MST6600_Fall2024/shapefiles/tl_2024_49_tract/tl_2024_49_tract.shp' 
  using driver `ESRI Shapefile'
Simple feature collection with 716 features and 13 fields
Geometry type: POLYGON
Dimension:     XY
Bounding box:  xmin: -114.0529 ymin: 36.99766 xmax: -109.0416 ymax: 42.0017
Geodetic CRS:  NAD83

Code

# Filter the shapefile to Salt Lake County (FIPS code: 035)
sl_county_sf <- ut_shapefile |>
  filter(COUNTYFP == "035")

sl_county_sf

Using `ggplot` with Shapefiles

Code

sl_county_sf |>
  ggplot() +
  geom_sf() +
  labs(
    title = "Salt Lake County Shapefile Plot",
    subtitle = "Census Tracts
    ") +
  theme_void()

Adding Data to the Shapefile

Download median household income data from the 5-Year American Community Survey (ACS)

Code

# Load required libraries
library(tidycensus)

# Set your Census API key
# You need to sign up for an API key at http://api.census.gov/data/key_signup.html
# census_api_key("YOUR_API_KEY_HERE", install = TRUE)

# Define parameters
year <- 2022  # Use the most recent year available
state <- "UT"
county <- "Salt Lake"

# Get median household income data
salt_lake_income <- get_acs(
  geography = "tract",
  variables = "B19013_001",  # Median household income variable
  state = state,
  county = county,
  year = year,
  survey = "acs5" # 5-year estimates
)

# View the data
salt_lake_income

Joining Two Data Frames: Example 1

Code

# Create two data frames
df1 <- tibble(id = c(1, 2, 3), value = c(10, 20, 30))
df2 <- tibble(id = c(1, 2, 3), value = c(100, 200, 300))

# Join the data frames
df_new <- df1 |>
  left_join(df2, by = "id")

df1

Code

df2

Code

df_new

Joining Two Data Frames: Example 2

Code

# Create two data frames with different IDs
df1 <- tibble(id = c(1, 2, 3), value = c(10, 20, 30))
df2 <- tibble(id = c(2, 3, 4), value = c(100, 200, 300))

# Join the data frames
df_new <- df1 |>
  left_join(df2, by = "id")

df1

Code

df2

Code

df_new

Joining Two Data Frames: Example 3

Code

# Create two data frames with different IDs
df1 <- tibble(id = c(1, 2, 3), value = c(10, 20, 30))
df2 <- tibble(id = c(2, 3, 4), value = c(100, 200, 300))

# Join the data frames
df_new <- df1 |>
  inner_join(df2, by = "id")

df1

Code

df2

Code

df_new

Joining Two Data Frames: Example 4

Code

# Create two data frames with different IDs
df1 <- tibble(id = c(1, 2, 3), value = c(10, 20, 30))
df2 <- tibble(id = c(2, 3, 4), value = c(100, 200, 300))

# Join the data frames
df_new <- df1 |>
  full_join(df2, by = "id")

df1

Code

df2

Code

df_new

Join the Data to the Shapefile

Code

sl_county_tracts_income <- sl_county_sf |>
  left_join(salt_lake_income, by = "GEOID")

sl_county_tracts_income

Create a Choropleth Map

Code

sl_county_tracts_income |>
  ggplot() +
  geom_sf(aes(fill = estimate)) +
  scale_fill_viridis_c(
    name = "Median Household\nIncome",
    label = scales::dollar,
    trans = "log10",
    breaks = c(30000, 60000, 120000, 240000),
  ) +
  theme_minimal() +
  labs(
    title = "Salt Lake County American Community Survey Data",
    subtitle = "2018 \u2013 2022; Table B19013_001"
    ) +
  theme_void()

Redlining (SLC)

Redlining Map of Salt Lake City, Utah

ca. 1933 - 1939

Maps Assignment

For this assignment, we will summarize the mean debt of graduate student loans per state and display the data on a map of the U.S.

You will need the College Scorecard dataset Most-Recent-Cohorts-Field-of-Study.csv and the Most-Recent-Cohorts-Institution.csv files.

We will go through the following steps in class:

Add directory information for each institution to the College Scorecard data frame
Clean the data frame by removing the “Privacy Supressed” fields and converting needed columns from text to numbers
Summarize the total debt by state
Decide what additional information is needed to join the summary data to the map data files
Create a choropleth map

Links:

College Scorecard Data

College Scorecard Documentation

Download the “Most Recent Data by Field of Study.zip” and the “Data Dictionary”

College Scorecard Datasets

Code

college_scorecard_field_of_study <- read_csv("datasets/College_Scorecard_Raw_data_06102024/Most-Recent-Cohorts-Field-of-Study.csv")

Code

dim(college_scorecard_field_of_study)

[1] 229188    174

Code

college_scorecard_institution <- read_csv("datasets/College_Scorecard_Raw_data_06102024/Most-Recent-Cohorts-Institution.csv")

Code

dim(college_scorecard_institution)

[1] 6484 3305

Maps

Libraries

Simple Maps (1/2)

Getting started with the maps package

Simple Maps (2/2)

Using Data from the maps Package in ggplot2 (1/3)

Create a data frame from the maps package databases

Using Data from the maps Package in ggplot2 (2/3)

Create a map of the US states using ggplot2

Using Data from the maps Package in ggplot2 (3/3)

Two solutions for scaling the map

1. Use the coord_fixed() function

2. Use the coord_map() function

Add Cities to the Map (1/4)

Create a data frame of cities in the US usign the us.cities data set

Add Cities to the Map (2/4)

Create a map of the US states with cities using ggplot2

Add Cities to the Map (3/4)

What’s a quick way to remove Alaska and Hawaii cities from the map?

Add Cities to the Map (4/4)

Add state capitals and the national capital as unique points

Map projections

Maps that Include Alaska and Hawaii (1/2)

Maps that Include Alaska and Hawaii (2/2)

the usmap::plot_usmap() function returns a ggplot object

Inspect the us_map() Data Frame

Note, we do not have a data frame with latitude and longitude columns

Adding Cities to the usmap Map

A convenience method called usmap_transform can be used to add cities (or other points) to the map.

Choropleth Maps (from Claude.ai)

what is the origin of the word choropleth?

Add a Choropleth Layer to the Map

State Population Data

Working with Shapefiles

US Census

Loaing Shapefiles

Using ggplot with Shapefiles

Adding Data to the Shapefile

Download median household income data from the 5-Year American Community Survey (ACS)

Joining Two Data Frames: Example 1

Joining Two Data Frames: Example 2

Joining Two Data Frames: Example 3

Joining Two Data Frames: Example 4

Join the Data to the Shapefile

Create a Choropleth Map

Redlining (SLC)

Maps Assignment

College Scorecard Datasets

Getting started with the `maps` package

Using Data from the `maps` Package in `ggplot2` (1/3)

Create a data frame from the `maps` package databases

Using Data from the `maps` Package in `ggplot2` (2/3)

Create a map of the US states using `ggplot2`

Using Data from the `maps` Package in `ggplot2` (3/3)

1. Use the `coord_fixed()` function

2. Use the `coord_map()` function

Create a data frame of cities in the US usign the `us.cities` data set

Create a map of the US states with cities using `ggplot2`

the `usmap::plot_usmap()` function returns a ggplot object

Inspect the `us_map()` Data Frame

Adding Cities to the `usmap` Map

A convenience method called `usmap_transform` can be used to add cities (or other points) to the map.

Using `ggplot` with Shapefiles